Skip to main content

Application of Artificial Intelligence Methods in Modeling Corrosion of Cement and Sulfur Concrete in Sewer Systems

Abstract

The aim of this study was to predict the corrosion of cement concrete (CC) and sulfur concrete (SC) in sewer systems using three artificial intelligence (AI)-based techniques: Adaptive NeuroFuzzy Inference System (ANFIS), Genetic Programming (GP), and Multi Expression Programming (MEP). Two sets of chemical experiments in acidic solutions and biological tests using Thiobacillus thiooxidans were conducted to investigate the corrosion in concrete samples. For both tests, weight loss was used as the indicator of corrosion. Time and pH values in chemical tests were selected as the input variables, while in biological experiments only time was selected as the input variable. In addition, compressive strength (CS), weight loss, water absorption, and hole index ratio (HIR) were selected as the output variables for both experiments. Based on the results obtained in both experiments, a database of 256 values was established to train and test the prediction models. According to the results, all AI-based models were able to predict the weight loss with reasonable accuracy. In contrast, the proposed MEP model outperformed other models with R2 of 92.1 (training) and 87.9 (testing) for CC and 99.5 (training) and 97.7 (testing) for SC, respectively. This study shows that machine learning-based models can help engineers estimate the corrosion of concrete pipes in sewer systems.

Highlights

• ANFIS, GP and MEP models were used to model corrosion in cement and sulfur concrete

• All AI-based methods were able to predict corrosion in both types of concrete

• MEP showed better performance than other methods in the prediction of corrosion

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Data availability

Data are not available on data repositories.

Code availability (software application or custom code)

Codes are available.

References

Download references

Acknowledgements

The authors gratefully thank Dr. Alireza Fallahpour for ANN consultations, and the manager and experts of Shahid Mahalati wastewater treatment plant, especially Eng. Jafarizadeh.

Funding

This project was funded by KNTU university.

Author information

Authors and Affiliations

Authors

Contributions

Conceptualization: Mohammad Reza Sabour, Ghorban Ali Dezvareh; Methodology: Ghorban Ali Dezvareh; Mohammad Reza Sabour; Formal analysis and investigation: Ghorban Ali Dezvareh; Kasra Pourrostami Niavol; Writing—original draft preparation: Kasra Pourrostami Niavol; Writing—review and editing: Kasra Pourrostami Niavol; Ghorban Ali Dezvareh; Supervision: Mohammad Reza Sabour.

Corresponding author

Correspondence to Kasra Pourrostami Niavol.

Ethics declarations

Conflicts of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Sabour, M.R., Dezvareh, G.A. & Niavol, K.P. Application of Artificial Intelligence Methods in Modeling Corrosion of Cement and Sulfur Concrete in Sewer Systems. Environ. Process. 8, 1601–1618 (2021). https://doi.org/10.1007/s40710-021-00542-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s40710-021-00542-y

Keywords

  • AI-based modeling
  • ANFIS
  • MEP
  • GP
  • Predictive models
  • Concrete corrosion
  • Sulfur concrete